eduKateSG · Why Science?
Read DNA modification sites with their supporting counts—while preserving coverage, percentages, call categories, strand, assay and uncertainty
Choose the closest route, then return to the full index whenever you need the wider evidence chain.
Science learning becomes powerful when students can tell a measured pattern from a model, a statistical decision from a mechanism, and a useful lead from a finished conclusion. UCSC describes bedMethyl as a BED9 extension for base-resolution DNA methylation data from bisulfite sequencing or direct modification detection such as long-read sequencing. Nine extra fields report valid coverage, percent modified, modified and canonical calls, other modifications, deletions, low-confidence calls, mismatches and reads with no modification call. bigMethyl is the indexed bed9+9 version built with bedToBigBed and an autoSql schema. These tracks make support visible, but a methylation percentage remains assay-, caller-, threshold- and sample-dependent. This guide supports Primary Science, PSLE Science, Secondary Science, O-Level Science and STEM education as a longform evidence-reading route; it is not medical advice, a laboratory protocol, an admission promise or a career guarantee.
Related eduKate reading: BED and genomic feature coordinates; bigBed indexed genome annotation tracks; bigWig indexed genomic signal tracks; Science Learning Hub; Education Hub; STEM Education; Careers by Subject Capability.
Primary and official sources checked on 12 October 2026: UCSC Genome Browser bedMethyl and bigMethyl Track Format, checked 12 October 2026; ENCODE bedMethyl format definition, checked 12 October 2026; UCSC Genome Browser bigBed Track Format, checked 12 October 2026; 2026 Singapore–Cambridge O-Level Biology syllabus; MOE G2/G3 Lower Secondary Science teaching and learning syllabus; 2026 MOE G2 Computing syllabus. Specifications, tools and recommendations can change, so preserve dated documentation, exact source data, coordinate rules and software versions.
A deeper reading route
bedMethyl keeps location, display and evidence counts in one record. The BED9 prefix places a base-resolution item in the browser. The nine added fields say how many reads had valid calls, what percentage were modified and how remaining observations were classified. This is more informative than a colour alone because readers can inspect denominator size and failure categories before trusting the percentage.
UCSC presents bedMethyl for methylation calls from bisulfite sequencing and direct long-read modification detection. Those technologies do not measure identical intermediate signals or share one universal error model. A track should therefore name assay, base-modification code, chemistry or platform, caller, probability threshold and sample context rather than presenting every red mark as the same kind of observation.
The BED9 prefix contains chromosome, start, end, name, score, strand, thick bounds and colour. In UCSC’s example, one-base intervals represent modification sites and colour runs from blue at zero percent modified to red at one hundred percent. The stored percentage and call counts are the evidence; colour is a visual encoding that can be changed without changing the underlying calls.
nValidCov is the count of reads with a valid modification call, and percMod is the percentage of valid calls that are modified. nMod and nCanon separate modified from canonical calls, while nOther records other modifications. Denominator rules matter: a reader should be able to verify whether nValidCov agrees with the included call categories under the producing specification.
nDelete, nFail, nDiff and nNoCall keep observations that do not simply join the modified-versus-canonical denominator. Deletions remove the reference base from that read; low-confidence calls fail the probability threshold; mismatches show another base; and no-call reads align with the expected canonical base but lack a modification call. Collapsing these categories into zero modification would bias the estimate.
The name field can distinguish methylation contexts or modification codes, and strand can matter for some analyses even though the example schema shown by UCSC labels the bigMethyl strand as unused. The documentation and data producer should be followed together. Never infer biological symmetry or strand neutrality solely from a placeholder character in a display schema.
bigMethyl uses a bigBed container with a documented bed9+9 declaration and bigMethyl.as schema. Input must be sorted and validated against assembly-matched chromosome sizes. The resulting binary file is hosted at a web-accessible URL for regional access. Indexing improves responsiveness but cannot correct a wrong reference base, sample swap or poorly calibrated caller.
Validation should recalculate percMod from documented count rules, check all counts are non-negative integers, ensure one-base or intended intervals fit the assembly, and test zero coverage without dividing by zero. Include all-modified, all-canonical, low-coverage, other-modification, deletion, failure, mismatch and no-call fixtures.
Cross-sample comparison requires more than aligning track colours. Coverage depth, cell mixture, assay efficiency, modification type and caller calibration can differ. Statistical analysis should model appropriate uncertainty and multiple testing. A ten-read site at fifty percent and a thousand-read site at fifty percent share a point estimate but not the same precision.
Methylation association is not automatically gene regulation or disease mechanism. Genomic context, cell type, developmental state and causal direction matter. Clinical or health conclusions need appropriate validated workflows and domain oversight. This educational guide is evidence-literacy material rather than medical interpretation or a diagnostic protocol.
For students, bedMethyl is like a classroom vote with an attendance sheet. The percentage is useful only after we know how many valid votes existed and how many responses were absent, unclear or in another category. The cheerful scientific habit is to look for the denominator before being impressed by the colour.
A reproducible package includes raw-data references, assay and caller versions, probability threshold, modification-code definitions, bedMethyl text, bigMethyl schema, assembly manifest, conversion commands, track settings, binary hash and count-recalculation tests. It keeps every displayed percentage traceable to counts and methods.
| Stage | Question | Evidence | Responsible output |
|---|---|---|---|
| Locate | Which base or interval is represented? | BED9 coordinates, name, strand and display fields | An assembly-specific site |
| Count | Which calls support the site? | Valid, modified, canonical and other-modification counts | A visible denominator and numerator |
| Qualify | What happened to remaining reads? | Deletion, low-confidence, mismatch and no-call counts | Explicit non-simple observations |
| Judge | What can percent modified support? | Assay, thresholds, coverage, samples and validation | A bounded modification estimate |
Inside this guide
1–12 · Foundations and design
- 1. Start with the modification question
- 2. Separate assay from format
- 3. Read chromosome and bounds
- 4. Read the modification name
- 5. Read score carefully
- 6. Read strand deliberately
- 7. Read thick bounds and colour
- 8. Read valid coverage
- 9. Read percent modified
- 10. Read modified calls
- 11. Read canonical calls
- 12. Read other modifications
13–24 · Models and estimation
- 13. Read deletion counts
- 14. Read low-confidence calls
- 15. Read base mismatches
- 16. Read no-call observations
- 17. Check count identities
- 18. Handle zero coverage
- 19. Respect low coverage
- 20. Prepare sorted input
- 21. Use the exact schema
- 22. Use assembly-matched sizes
- 23. Build as bed9+9
- 24. Host for ranged access
25–36 · Diagnostics and interpretation
- 25. Round-trip all eighteen fields
- 26. Protect sensitive epigenomic data
- 27. Separate modification from mechanism
- 28. Connect to Primary Science
- 29. Connect to Secondary Science
- 30. Connect to mathematics
- 31. Connect to computing
- 32. Connect to pathways
- 33. Did you know?
- 34. Write a reader-ready modification note
- 35. Build a validation ladder
- 36. Finish with a reproducible bundle
Section 1 of 36
1. Start with the modification question
bedMethyl stores site-level DNA modification estimates and support categories.
Working checkpoint — bedMethyl: Name sample, assay, modification and intended comparison. Keep the independent sample, design, annotation and software version beside this decision so the evidence can be reconstructed.
For a classroom model, use invented counts and clearly label the activity as an analogy rather than a sequencing experiment.
Section 2 of 36
2. Separate assay from format
The record can hold calls from different technologies whose measurement models differ.
Working checkpoint — bigMethyl: Publish platform, chemistry and caller. Read the numerical result together with sample-level plots, effect size and a sensitivity check capable of weakening the headline.
For future study, connect the idea to biology, mathematics and computing while checking current official pathway information separately.
Section 3 of 36
3. Read chromosome and bounds
The BED prefix locates the item on an exact reference assembly.
Working checkpoint — DNA methylation: Verify reference base and legal interval. Preserve the complete tested family and every excluded feature; a polished shortlist is not a substitute for provenance.
For scientific writing, prefer “supports”, “is associated with” and “is predicted to” when stronger causal verbs are not justified.
Section 4 of 36
4. Read the modification name
name can identify methylation context or modification type under producer conventions.
Working checkpoint — base modification: Provide a codebook. Separate what the model estimates from the biological story proposed afterward, then name the extra evidence that story would require.
For students, the transferable habit is to ask: What was measured? Compared with what? What alternative explanation remains?
Section 5 of 36
5. Read score carefully
score is a zero-to-one-thousand display field under BED rules.
Working checkpoint — coverage: Do not substitute it for coverage or percentage. Treat identifiers, denominators and factor levels as scientific quantities: a silent mismatch can reverse an otherwise correct calculation.
For parents, a calm checkpoint is to ask how many independent samples support the claim and which assumption matters most.
Section 6 of 36
6. Read strand deliberately
strand semantics depend on the producer and displayed schema.
Working checkpoint — percent modified: State whether strand is measured, collapsed or unused. Keep the independent sample, design, annotation and software version beside this decision so the evidence can be reconstructed.
For a classroom model, use invented counts and clearly label the activity as an analogy rather than a sequencing experiment.
Section 7 of 36
7. Read thick bounds and colour
thickStart, thickEnd and colour support display.
Working checkpoint — bedMethyl: Keep colour legend accessible and documented. Read the numerical result together with sample-level plots, effect size and a sensitivity check capable of weakening the headline.
For future study, connect the idea to biology, mathematics and computing while checking current official pathway information separately.
Section 8 of 36
8. Read valid coverage
nValidCov counts reads with a valid modification call.
Working checkpoint — bigMethyl: Use it as the documented percentage denominator. Preserve the complete tested family and every excluded feature; a polished shortlist is not a substitute for provenance.
For scientific writing, prefer “supports”, “is associated with” and “is predicted to” when stronger causal verbs are not justified.
Section 9 of 36
9. Read percent modified
percMod reports the percentage of valid calls that are modified.
Working checkpoint — DNA methylation: Recalculate it from source counts. Separate what the model estimates from the biological story proposed afterward, then name the extra evidence that story would require.
For students, the transferable habit is to ask: What was measured? Compared with what? What alternative explanation remains?
Section 10 of 36
10. Read modified calls
nMod counts calls assigned to the represented modification.
Working checkpoint — base modification: Keep modification identity explicit. Treat identifiers, denominators and factor levels as scientific quantities: a silent mismatch can reverse an otherwise correct calculation.
For parents, a calm checkpoint is to ask how many independent samples support the claim and which assumption matters most.
Section 11 of 36
11. Read canonical calls
nCanon counts calls assigned to the canonical base.
Working checkpoint — coverage: Do not treat uncalled reads as canonical. Keep the independent sample, design, annotation and software version beside this decision so the evidence can be reconstructed.
For a classroom model, use invented counts and clearly label the activity as an analogy rather than a sequencing experiment.
Section 12 of 36
12. Read other modifications
nOther separates calls for other modification types.
Working checkpoint — percent modified: Preserve them rather than merging into one negative class. Read the numerical result together with sample-level plots, effect size and a sensitivity check capable of weakening the headline.
For future study, connect the idea to biology, mathematics and computing while checking current official pathway information separately.
Section 13 of 36
13. Read deletion counts
nDelete records reads with a deletion at the reference position.
Working checkpoint — bedMethyl: Exclude according to documented denominator rules. Preserve the complete tested family and every excluded feature; a polished shortlist is not a substitute for provenance.
For scientific writing, prefer “supports”, “is associated with” and “is predicted to” when stronger causal verbs are not justified.
Section 14 of 36
14. Read low-confidence calls
nFail records calls below the probability threshold.
Working checkpoint — bigMethyl: Publish the threshold and calibration method. Separate what the model estimates from the biological story proposed afterward, then name the extra evidence that story would require.
For students, the transferable habit is to ask: What was measured? Compared with what? What alternative explanation remains?
Section 15 of 36
15. Read base mismatches
nDiff records reads with a base other than the canonical base for this modification.
Working checkpoint — DNA methylation: Inspect variant and alignment context. Treat identifiers, denominators and factor levels as scientific quantities: a silent mismatch can reverse an otherwise correct calculation.
For parents, a calm checkpoint is to ask how many independent samples support the claim and which assumption matters most.
Section 16 of 36
16. Read no-call observations
nNoCall records aligned reads with the expected base but no modification call.
Working checkpoint — base modification: Keep missingness separate from zero. Keep the independent sample, design, annotation and software version beside this decision so the evidence can be reconstructed.
For a classroom model, use invented counts and clearly label the activity as an analogy rather than a sequencing experiment.
Section 17 of 36
17. Check count identities
Producer rules should explain how valid and non-valid categories relate.
Working checkpoint — coverage: Flag impossible totals and negative counts. Read the numerical result together with sample-level plots, effect size and a sensitivity check capable of weakening the headline.
For future study, connect the idea to biology, mathematics and computing while checking current official pathway information separately.
Section 18 of 36
18. Handle zero coverage
A site without valid calls cannot support an ordinary percentage estimate.
Working checkpoint — percent modified: Use explicit missingness, not zero modification. Preserve the complete tested family and every excluded feature; a polished shortlist is not a substitute for provenance.
For scientific writing, prefer “supports”, “is associated with” and “is predicted to” when stronger causal verbs are not justified.
Section 19 of 36
19. Respect low coverage
A percentage based on very few valid calls has substantial uncertainty.
Working checkpoint — bedMethyl: Show counts beside percentages. Separate what the model estimates from the biological story proposed afterward, then name the extra evidence that story would require.
For students, the transferable habit is to ask: What was measured? Compared with what? What alternative explanation remains?
Section 20 of 36
20. Prepare sorted input
bedMethyl rows must be ordered for bigBed conversion.
Working checkpoint — bigMethyl: Verify chromosome and numeric start order. Treat identifiers, denominators and factor levels as scientific quantities: a silent mismatch can reverse an otherwise correct calculation.
For parents, a calm checkpoint is to ask how many independent samples support the claim and which assumption matters most.
Section 21 of 36
21. Use the exact schema
bigMethyl.as defines BED9 plus nine evidence fields.
Working checkpoint — DNA methylation: Hash the schema with the output. Keep the independent sample, design, annotation and software version beside this decision so the evidence can be reconstructed.
For a classroom model, use invented counts and clearly label the activity as an analogy rather than a sequencing experiment.
Section 22 of 36
22. Use assembly-matched sizes
chrom.sizes fixes legal sequence names and coordinate limits.
Working checkpoint — base modification: Reject aliases and overflow. Read the numerical result together with sample-level plots, effect size and a sensitivity check capable of weakening the headline.
For future study, connect the idea to biology, mathematics and computing while checking current official pathway information separately.
Section 23 of 36
23. Build as bed9+9
bedToBigBed must receive the documented declaration and schema.
Working checkpoint — coverage: Record full command and version. Preserve the complete tested family and every excluded feature; a polished shortlist is not a substitute for provenance.
For scientific writing, prefer “supports”, “is associated with” and “is predicted to” when stronger causal verbs are not justified.
Section 24 of 36
24. Host for ranged access
bigMethyl supports regional retrieval from a web-accessible URL.
Working checkpoint — percent modified: Test anonymous access and byte ranges. Separate what the model estimates from the biological story proposed afterward, then name the extra evidence that story would require.
For students, the transferable habit is to ask: What was measured? Compared with what? What alternative explanation remains?
Section 25 of 36
25. Round-trip all eighteen fields
Extracted rows should reproduce every coordinate, count and percentage.
Working checkpoint — bedMethyl: Compare values and field types. Treat identifiers, denominators and factor levels as scientific quantities: a silent mismatch can reverse an otherwise correct calculation.
For parents, a calm checkpoint is to ask how many independent samples support the claim and which assumption matters most.
Section 26 of 36
26. Protect sensitive epigenomic data
Human modification profiles can be sensitive and re-identifiable.
Working checkpoint — bigMethyl: Apply governance before public hosting. Keep the independent sample, design, annotation and software version beside this decision so the evidence can be reconstructed.
For a classroom model, use invented counts and clearly label the activity as an analogy rather than a sequencing experiment.
Section 27 of 36
27. Separate modification from mechanism
A site-level percentage does not by itself prove regulation, phenotype or disease causality.
Working checkpoint — DNA methylation: Require independent evidence for stronger claims. Read the numerical result together with sample-level plots, effect size and a sensitivity check capable of weakening the headline.
For future study, connect the idea to biology, mathematics and computing while checking current official pathway information separately.
Section 28 of 36
28. Connect to Primary Science
A voting example shows why a percentage needs a valid-response denominator and missing categories.
Working checkpoint — base modification: Use fictional non-health data. Preserve the complete tested family and every excluded feature; a polished shortlist is not a substitute for provenance.
For scientific writing, prefer “supports”, “is associated with” and “is predicted to” when stronger causal verbs are not justified.
Section 29 of 36
29. Connect to Secondary Science
Modification tracks illustrate measurement, classification, controls and uncertainty in genetics.
Working checkpoint — coverage: Separate DNA sequence from chemical modification. Separate what the model estimates from the biological story proposed afterward, then name the extra evidence that story would require.
For students, the transferable habit is to ask: What was measured? Compared with what? What alternative explanation remains?
Section 30 of 36
30. Connect to mathematics
Percentages, denominators, coverage and sampling uncertainty create meaningful quantitative practice.
Working checkpoint — percent modified: Compare equal percentages at different depths. Treat identifiers, denominators and factor levels as scientific quantities: a silent mismatch can reverse an otherwise correct calculation.
For parents, a calm checkpoint is to ask how many independent samples support the claim and which assumption matters most.
Section 31 of 36
31. Connect to computing
Typed counts, binary indexes and validation rules show how evidence becomes a data product.
Working checkpoint — bedMethyl: Test zero and missing values. Keep the independent sample, design, annotation and software version beside this decision so the evidence can be reconstructed.
For a classroom model, use invented counts and clearly label the activity as an analogy rather than a sequencing experiment.
Section 32 of 36
32. Connect to pathways
Epigenomics joins biology, chemistry, statistics, sequencing and software and can illuminate STEM pathways.
Working checkpoint — bigMethyl: Verify current study routes separately. Read the numerical result together with sample-level plots, effect size and a sensitivity check capable of weakening the headline.
For future study, connect the idea to biology, mathematics and computing while checking current official pathway information separately.
Section 33 of 36
33. Did you know?
bedMethyl keeps nine call-support fields after BED9, including deletions, failed calls, mismatches and no-calls.
Working checkpoint — DNA methylation: Inspect these before trusting percent modified. Preserve the complete tested family and every excluded feature; a polished shortlist is not a substitute for provenance.
For scientific writing, prefer “supports”, “is associated with” and “is predicted to” when stronger causal verbs are not justified.
Section 34 of 36
34. Write a reader-ready modification note
Name assay, caller, thresholds, modification code, assembly, counts, schema and validation.
Working checkpoint — base modification: Keep it beside the track. Separate what the model estimates from the biological story proposed afterward, then name the extra evidence that story would require.
For students, the transferable habit is to ask: What was measured? Compared with what? What alternative explanation remains?
Section 35 of 36
35. Build a validation ladder
Check coordinates, base, strand, counts, denominator, percentage, conversion, hosting and extraction.
Working checkpoint — coverage: Stop at the first impossible total. Treat identifiers, denominators and factor levels as scientific quantities: a silent mismatch can reverse an otherwise correct calculation.
For parents, a calm checkpoint is to ask how many independent samples support the claim and which assumption matters most.
Section 36 of 36
36. Finish with a reproducible bundle
Deliver inputs, calls, codebooks, schemas, manifests, commands, hashes and bounded claims.
Working checkpoint — percent modified: Make every percentage traceable to counts. Keep the independent sample, design, annotation and software version beside this decision so the evidence can be reconstructed.
For a classroom model, use invented counts and clearly label the activity as an analogy rather than a sequencing experiment.
Contents · Previous section · Continue to the Science Learning Hub